Prepared by Tara Jagadeesh
Image by Bonanza Creek LTER
February 2020

Project Summary

This project examines characteristics of snowshoe hares in the Bonanza Creek Experimental Forest (BCEF). First, we consider the gender distribution of population over time by finding the proportion of females caught in traps. We then examine how average weight differs between female and male hares at three different study sites (bonbs, bonmat, bonrip). Data used in this projected was collected from 1999 to 2012 by Kielland et al. (2017) using the capture recapture method.

Study Area and Species of Interest


Map of Bonanza Creek Experimental Forest        Snowshoe Hare

         Figure 1: Map of Bonanza Creek Experimental Forest          Figure 2: Snowshoe Hare


Data Analysis

#Load packages
library(tidyverse)
library(janitor)
library(lubridate)
library(kableExtra)

#Read in snowshoe hare data
hares <- read_csv("showshoe_lter.csv")
#Clean up data
hares_clean <- hares %>% 
  #Convert all entries to lowercase
  mutate(site = str_to_lower(grid), trap = str_to_lower(trap), sex = str_to_lower(sex), 
         age = str_to_lower(age), animal_id = b_key, study = str_to_lower(study)) %>% 
  #Add a column for year of observation only
  mutate(date = mdy(date), year = year(date)) %>% 
  select(date, year, site, trap, animal_id, sex, age, weight, hindft, session_id, 
         study, notes) #Keep columns of interest

#Create a simple dataframe with variables of interest
hares_weight <- hares_clean %>% 
  select(year, site, sex, weight, age) %>% 
  drop_na() %>% 
  #Include only rows which contain "m" or "f" in the sex column
  filter(sex %in% c("m", "f")) %>% 
  filter(age == "a") #Include only identified adult hares
#Calculate the proportion of female snowshoe hares trapped each year
sex_summary <- hares_weight %>% 
  group_by(year, sex) %>%  
  #Count the number of males and females trapped each year
  tally() %>%  
  #Sum males and females to find the total number of trappings each year
  mutate(sum = sum(n)) %>% 
  #Create a new column with the proportion of females
  mutate(ratio = n/sum) %>% 
  #Filter for females only and remove columns with insufficient data
  filter(year != 2002, year != 2003, sex == "f") 

#Make a graph comparing the proportion of female snowshoe hares trapped each year
sex_graph <- ggplot(data = sex_summary, aes(x= year, y = ratio)) +
  geom_col(fill = "darkslateblue") +
  scale_y_continuous(expand = c(0,0), limits = c(0, 1.1)) +
  scale_x_continuous(expand = c(0,0), breaks=seq(1999, 2012))+
  labs(x = "\n Year", y = "Proportion of females \n") +
  geom_text(aes(label = paste("n =",sum)), vjust=-0.5, size = 2.5) +
  theme_classic() 
sex_graph

Figure 1. Proportion of female snowshoe hares captured in traps annually (1999 2012). Due to insufficient data 2002, 2003, 2004 and were omitted. Sample size is given by n. Proportion of females varies by year, with most years exceeding 50% female. Data source: Kielland et al. (2017)


#Calculate summary statistics for weight 
hares_summary <- hares_weight %>% 
  group_by(site, sex) %>%  
  summarize( 
    n = length(weight),
    weight_avg = mean(weight),
    weight_sd = sd(weight),
    weight_min = min(weight),
    weight_max = max(weight)) %>% 
  mutate(sex = str_to_upper(sex)) %>% 
  select(sex, n, weight_avg, weight_sd, weight_min, weight_max)

#Remove the "site" column for easier table grouping
hares_summary2 <- hares_summary %>% 
  ungroup() %>% 
 select(-site)
#Create a table of the summary statistics 
table <- kable(hares_summary2, booktabs = T, col.names = c("Sex", "N", "Mean", "S.D.", "Min", "Max"), digits = 0, 
      caption = "Table 1: Summary statistics for snowshoe hare weight by sex at three
      different trapping sites (bonbs, bonmst, bonrip). Each site is located within the boundaries of the Bonanza Creek Experimental Forest. Data is averaged from 1999 to 2012, and units for weight are in grams.\n") %>% 
  kable_styling(bootstrap_options = c("striped"), full_width = T) %>% 
  pack_rows("Bonbs (n = 68)", 1, 2) %>% #Add group headings that correspond to site
  pack_rows("Bonmat (n = 120)", 3, 4) %>%
  pack_rows("Bonrip (n = 494)", 5, 6)

table
Table 1: Summary statistics for snowshoe hare weight by sex at three different trapping sites (bonbs, bonmst, bonrip). Each site is located within the boundaries of the Bonanza Creek Experimental Forest. Data is averaged from 1999 to 2012, and units for weight are in grams.
Sex N Mean S.D. Min Max
Bonbs (n = 68)
F 24 1602 172 1330 1960
M 44 1480 164 1225 1780
Bonmat (n = 120)
F 51 1492 263 800 1930
M 69 1478 196 960 1800
Bonrip (n = 494)
F 309 1520 264 720 2365
M 185 1451 209 730 2050